github/ColossalAI
1
0
Fork 0
mirror of https://github.com/hpcaitech/ColossalAI.git synced 2025-09-04 18:40:28 +00:00
Code Issues Packages Projects Releases Wiki Activity

fix schedualing for multi-node training

Browse Source
This commit is contained in:
YeAnbang
2025-05-02 19:45:07 +08:00
parent d06042b434
commit 7d658402da
7 changed files with 124 additions and 38 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
applications/ColossalChat/coati/distributed/consumer.py
View File

@@ -115,10 +115,10 @@ class BaseConsumer:
eval_statistics = None
eval_global_step = None
for r in range(self.num_producers):
print(f"[T{dist.get_rank()}] Recv eval result episode {episode} step {step} from {r}")
local_eval_result = ray_broadcast_tensor_dict(
None, src=0, device=self.device, group_name=f"sync_data_{r}"
)
print(f"[T{dist.get_rank()}] Recv eval result episode {episode} step {step} from {r}")
assert "consumer_global_step" in local_eval_result
eval_global_step = local_eval_result.pop("consumer_global_step").item()
if eval_statistics is None: